The last five Housing surveys (1984-2002) confirm the role of the factors that usually explain residential mobility (age, income, level of education, family structure and housing characteristics). They can also be used to refine certain analyses and correct certain preconceived ideas. Among the household and housing characteristics, the importance of age and occupational standing are corroborated. Family events prove to be more decisive factors for residential mobility than employment considerations. Employment only plays the lead role in cases of long-distance mobility, alongside qualifications, age, income and household size. Short-distance mobility remains in the majority even though it is posting a relative downturn. The causes of mobility have changed little overall in the last twenty years. However, other things being equal, low rental public housing tenants were more mobile than private sector tenants in 1984 whereas the reverse was true in 2002. Lastly, the overall economic situation and the state of the employment market in particular strongly influence residential mobility, especially tenant mobility. Changes to household composition and employment conditions, factors that affect mobility the most, were also found the least frequently in 2002: formation and separation of couples, change of establishments and, lastly, births.